/**
 * Copyright (C) 2010-2015 The Roslin Institute <contact andy.law@roslin.ed.ac.uk>
 *
 * This file is part of JEnsembl: a Java API to Ensembl data sources developed by the
 * Bioinformatics Group at The Roslin Institute, The Royal (Dick) School of
 * Veterinary Studies, University of Edinburgh.
 *
 * Project hosted at: http://jensembl.sourceforge.net
 *
 * This is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License (version 3) as published by
 * the Free Software Foundation.
 *
 * This software is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * in this software distribution. If not, see: http://opensource.org/licenses/gpl-3.0.html
 */
package uk.ac.roslin.ensembl.demo;


import java.io.File;
import java.nio.charset.Charset;
import java.nio.file.Files;
import java.nio.file.Paths;
import java.util.List;
import java.util.TreeMap;
import uk.ac.roslin.ensembl.config.DBConnection.DataSource;
import uk.ac.roslin.ensembl.config.EnsemblDBType;
import uk.ac.roslin.ensembl.dao.database.DBDatabase;
import uk.ac.roslin.ensembl.dao.database.DBRegistry;
import uk.ac.roslin.ensembl.dao.database.DBSingleSpeciesCoreDatabase;
import uk.ac.roslin.ensembl.dao.database.DBSpecies;
import uk.ac.roslin.ensembl.dao.database.factory.DBDAOSingleSpeciesCoreFactory;
import uk.ac.roslin.ensembl.datasourceaware.core.DAChromosome;
import uk.ac.roslin.ensembl.datasourceaware.core.DAGene;
import uk.ac.roslin.ensembl.model.core.Chromosome;
import uk.ac.roslin.ensembl.model.core.Gene;
import uk.ac.roslin.ensembl.model.database.Database;
import uk.ac.roslin.ensembl.model.database.SingleSpeciesCoreDatabase;


public class EnsemblConnection {

    //Demonstrating typical connection to the Ensembl (Vertebrate) datasource, 
    //Registry autoconfiguration and basic data retrieval functions.
    public  static void main(String[] args) throws Exception {


        //we specify the Datasource for connection, and initialize the registry
        //this reads all the databases and species from the source
        
        long start = System.currentTimeMillis();
        DBRegistry ensembldbRegistry =DBRegistry.createRegistryForDataSource(DataSource.ENSEMBLDB);      
        System.out.println("Initialising Registry took: " + (System.currentTimeMillis() - start) / 1000 + " seconds");
        
        System.out.println(ensembldbRegistry.getVersionReport());
        
        //print a detailed report on this 
        System.out.println("ensembldb registry report: ");
        System.out.println("************************** ");
        File f = ensembldbRegistry.getRegistryReport();
        List<String> lines = Files.readAllLines(Paths.get(f.getCanonicalPath()), Charset.forName("UTF-8")); 
        for (String l:lines) {
            System.out.println(l);
        }
        
        
        //just print details on species present
        System.out.println("species count : "+ensembldbRegistry.getSpecies().size());
        int i = 1;
        for (DBSpecies s: ensembldbRegistry.getSpecies()) {
            System.out.println(i++ +": "+s.getSpeciesBinomial());
            for(DBDatabase d : s.getDatabasesByType(EnsemblDBType.core)) {
                System.out.println(d.getdBName()+ " : "+d.getBuild());
            }
        }
        
        
        
        //get a species
        DBSpecies human = ensembldbRegistry.getSpeciesByAlias("human");
        
                //you can add aliases to the hashmap read from the datasource meta tables...
        try {
            ensembldbRegistry.addSessionAlias("homo_sapiens", "Bob");
            System.out.println("added alias Bob ok");
        } catch (Exception exception) {
            System.out.println("failed to add alias Bob for homo_sapiens");
        }

        // aliases must be to the ensembl database style name
        try {
            ensembldbRegistry.addSessionAlias("homo sapiens", "Norman");
            System.out.println("added alias Norman ok - but shouldnt have");
        } catch (Exception exception) {
            System.out.println("failed to add alias Norman - this is correct behaviour");
            System.out.println("The correct database style name for humans is: "
                    + human.getDatabaseStyleName());
        }
        
        
        //get a chromosome by name - the current release version
        DAChromosome ch20 = human.getChromosomeByName("20");
        
        //get a chromosome by name and release version 
        DAChromosome ch20_65 = human.getChromosomeByName("20", "65");
        
        //get all the chromosomes at once
        //Note for humans this also get the patches
        TreeMap<String, DAChromosome> chrs = human.getChromosomes();

        for(String s: chrs.keySet()) {
               System.out.println("Human Chromosome: "+s+" length: "+chrs.get(s).getLength());
         }
        
        //get some genes on a chromosome
        List<DAGene> genesOnRegion = ch20.getGenesOnRegion(1,1000000);
        
        // get all the genes on a chromosome
        List<DAGene> allGenes = ch20.getGenesOnRegion(ch20.getBioBegin(), ch20.getBioEnd());
        //or
        List<DAGene> allGenes2 = ch20.getGenesOnRegion(1, ch20.getLength());
        
        System.out.println("count of genes on human chromosome 1: "+allGenes.size());
        
        //you can get and query Database objects - but this is not typically necessary
        //as these data access methods are wrapped by the datasource-aware layer
        SingleSpeciesCoreDatabase db = (SingleSpeciesCoreDatabase) ensembldbRegistry.getDatabase("human");

        System.out.println("current human top level: "+db.getTopLevelCoordSystem().getId());
        System.out.println("current human chromosome level: "+db.getChromosomeLevelCoordSystem().getId());
        System.out.println("current human sequence level: "+db.getSequenceLevelCoordSystem().getId());

        //you can also get and use the factory object for a database that is responsible 
        //for correctly configuring the dataaccess queries


        
        DBDAOSingleSpeciesCoreFactory fact = ((DBSingleSpeciesCoreDatabase) ensembldbRegistry.getDatabase("chicken")).getCoreFactory();
        DBDAOSingleSpeciesCoreFactory fact60 = ((DBSingleSpeciesCoreDatabase) ensembldbRegistry.getDatabase("chicken","60")).getCoreFactory();
        DBDAOSingleSpeciesCoreFactory fact61 = ((DBSingleSpeciesCoreDatabase) ensembldbRegistry.getDatabase("chicken","61")).getCoreFactory();

        DAChromosome chr = fact.getChromosomeDAO().getChromosomeByName("25");
        DAChromosome chr60 = fact60.getChromosomeDAO().getChromosomeByName("25");
        DAChromosome chr61 = fact61.getChromosomeDAO().getChromosomeByName("25");

        System.out.println("Chromosome:\n\tName: "+chr.getChromosomeName()+ " db version-"+chr.getDBVersion());
        System.out.println("\tLength: "+chr.getDBSeqLength());
        System.out.println("\tcoord sys ID: "+chr.getCoordSystem().getId());
        System.out.println("\tseq level coord sys ID: "+((DBSingleSpeciesCoreDatabase) chr.getDaoFactory().getDatabase()).getSequenceLevelCoordSystem().getId());
        System.out.println("\tseq region ID :"+chr.getId());
        System.out.println("\tseq type :"+chr.getType().toString());

        System.out.println("Chromosome:\n\tName: "+chr60.getChromosomeName()+" db version-"+chr60.getDBVersion());
        System.out.println("\tLength: "+chr60.getDBSeqLength());
        System.out.println("\tcoord sys ID: "+chr60.getCoordSystem().getId());
        System.out.println("\tseq level coord sys ID: "+((DBSingleSpeciesCoreDatabase) chr60.getDaoFactory().getDatabase()).getSequenceLevelCoordSystem().getId());
        System.out.println("\tseq region ID :"+chr60.getId());
        System.out.println("\tseq type :"+chr60.getType().toString());

        System.out.println("Chromosome:\n\tName: "+chr61.getChromosomeName()+ " db version-"+chr61.getDBVersion());
        System.out.println("\tLength: "+chr61.getDBSeqLength());
        System.out.println("\tcoord sys ID: "+chr61.getCoordSystem().getId());
        System.out.println("\tseq level coord sys ID: "+ ((DBSingleSpeciesCoreDatabase) chr61.getDaoFactory().getDatabase()).getSequenceLevelCoordSystem().getId());
        System.out.println("\tseq region ID :"+chr61.getId());
        System.out.println("\tseq type :"+chr61.getType().toString());

        System.out.println("\n\n*****************************\n* COMPLETED FUNCTIONAL TEST *\n*****************************\n");

        
        
        System.out.println("Mapping all versions of ENSG00000153551");
        
        StringBuilder sb = new StringBuilder();
        start = System.currentTimeMillis();
        
        for (Database d : human.getDatabasesByType(EnsemblDBType.core)) {
            sb.append("Ensembl Release:" + d.getSchemaVersion() + " Assembly:" + d.getBuild());
            Gene g2 = human.getGeneByStableID("ENSG00000153551", d.getSchemaVersion());
            if (g2 != null) {
                sb.append("\tChromosome:" + ((Chromosome) g2.getChromosomeMapping().getTarget()).getName()
                        + " " + g2.getChromosomeMapping().getTargetCoordinates().toString());
            } else {
                sb.append("\tNot Present");
            }
            sb.append("\n");
        }

       System.out.println("Mapping all versions of ENSG00000153551 took: " + (System.currentTimeMillis() - start) / 1000 + " seconds");
       System.out.println(sb.toString());
    }

}